Estimation of the recharge area contributing water to a pumped well in a glacial-drift, river-valley aquifer

Abstract

The highly permeable, unconfined, glacial-drift aquifers that occupy most New England river valleys constitute the principal source of drinking water for many of the communities that obtain part or all of their public water supply from ground water. Recent events have shown that these aquifers are highly susceptible to contamination that results from a number of sources, such as seepage from wastewater lagoons, leaking petroleum-product storage tanks, and road salting. To protect the quality of water pumped from supply wells in these aquifers, it is necessary to ensure that potentially harmful contaminants do not enter the ground in the area that contributes water to the well. A high degree of protection can be achieved through the application of appropriate land-use controls within the contributing area. However, the contributing areas for most supply wells are not known. This report describes the factors that affect the size and shape of contributing areas to public supply wells and evaluates several methods that may be used to delineate contributing areas of wells in glacial-drift, river-valley aquifers. Analytical, two-dimensional numerical, and three-dimensional numerical models were used to delineate contributing areas. These methods of analysis were compared by applying them to a hypothetical aquifer having the dimensions and geometry of a typical glacial-drift, river-valley aquifer. In the model analyses, factors that control the size and shape of a contributing area were varied over ranges of values common to glacial-drift aquifers in New England. The controlling factors include the rate of well discharge, rate of recharge to the aquifer from precipitation and from adjacent till and bedrock uplands, distance of a pumping well from a stream or other potential source of induced recharge, degree of hydraulic connection of the aquifer with a stream, horizontal hydraulic conductivity of the aquifer, ratio of horizontal to vertical hydraulic conductivity, and degree of well penetration. Analytical methods proved easiest to apply but gave results that are considered to be less accurate than those obtainable by means of numerical-model analysis. Numerical models have the capability to more closely reflect the variable geohydrologic conditions typical of glacial-drift valley aquifers. For average conditions in the hypothetical aquifer, the analytical method predicts a contributing area limited to the well side of the river because a constant-head boundary simulated by image wells is used in the analytical model. For typical glacial-drift, river-valley aquifers, this simulation is unrealistic because drawdowns, caused by a pumping well, and the contributing area of the well can extend beneath and beyond a river or stream. A wide range of hydrologic conditions was simulated by using the two-dimensional numerical model. The resulting contributing area for a well pumped at 1.0 million gallons per day--a common pumping rate--ranged from about 0.9 to 1.8 square miles. Model analyses also show that the contributing area of pumped wells may be expected to extend to the opposite side of the river and to include significant areas of till uplands adjacent to the aquifer on both sides of the valley. Simulations done with the three-dimensional model allow a full three-dimensional delineation of the zone of contribution for a pumped well. For the relatively thin (100 feet or less) unconfined aquifers considered in this analysis, the three-dimensional model showed that the zone of contribution extended throughout the entire saturated thickness of aquifer; therefore, the two-dimensional simulations were considered adequate for delineating contributing areas in this particular hydrologic setting. For thicker aquifers, especially those having partially penetrating wells, three-dimensional models are preferable. Values for several of the factors that affect the size and shape of contributing recharge areas cannot be det